A Study On Visualization Analysis Of Space-time Series

Spatial and temporal sequence data exist in many fields such as meteorology,finance,transportation and so on.It is important to analyze the data of spatio-temporal sequence data,such as real-time monitoring and calculation of flight data analysis can help the detection and prediction of flight anomalies.However,due to the complex characteristics of space-time series data,such as massive,high-dimensional and dynamic,it brings great difficulty to its reasonable and effective analysis.So it is very important to study the analysis method of the spatiotemporal sequence data.In this paper,we focus on the visualization analysis of spatio-temporal data.Visualization analysis is the most intuitive way to display the data.It can greatly understand the large amount of data,and it is very helpful for the mining of the potential relationship and the rule of data.In this paper,we present a variety of visualization methods for displaying multidimensional attributes of spatial and temporal data,displaying data clustering procedure,displaying clustering results and important interactive operations.Moreover,making full use of the advantages of three-dimensional space and good interactive operations,this visualization tool contributes to effectively understanding and mastering the information bottleneck clustering algorithm.A visualization tool for 3D spatiotemporal data analysis was designed using unity3 D.In this tool,we focus on the visualization of trajectory attributes of pedestrians,the visualization of clustering process of information bottleneck algorithm,and the visualization of clustering results.In addition,this paper presents a novel dynamic process visualization design framework,especially for the each iterative step of the hierarchical clustering.Finally,we demonstrate the effectiveness of the information bottleneck clustering algorithm through testing a large number of simulated experimental data sets and real datasets,the superiority of the information bottleneck clustering algorithm and the actuality of the visualization tool through elaborating practical visualization examples.